Research map / Label-efficient and robust learning / OOD and anomaly detection
FDP: A Frequency-Decomposition Preprocessing Pipeline for Unsupervised Anomaly Detection in Brain MRI
Hao Li, Zhenfeng Zhuang, Jingyu Lin, Yu Liu, Yifei Chen, Qiong Peng, Lequan Yu, Liansheng Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add FDP (frequency decomposition preprocessing) | LDM · BraTS20 UAD | DICE / AUPRC: DICE 35.02 -> 52.66; AUPRC 30.75 -> 51.67 | Improved |
| add FDP | seven UAD baselines (VAE, LDM, AnoDDPM, F-AnoGAN, DAE simplex/coarse, pDDPM) · BraTS20 UAD | DICE / AUPRC: DICE gains from +6.16 to +17.63; AUPRC at least +5.36 | Improved |
| add FDP | UAD baselines · BraTS20 UAD | AUROC: VAE 94.46 -> 92.16 (-2.30); DAE (coarse) -1.76; LDM 91.62 -> 93.12 (+1.50) | Mixed |
| add FDP to LDM | LDM · BraTS21, MSLUB, MSSEG-2 | DICE: 29.53 -> 45.06; 8.35 -> 13.06; 20.15 -> 34.63 | Improved |
| combine FRM and HFSup (ablation) | FDP + LDM · BraTS20 UAD | DICE: 35.02 (neither) -> 52.66 (both); single-module variants 42.18 and 50.00 | Improved |
| set FRM high-pass threshold to 0.10 instead of 0.01 (ablation) | FDP + LDM · BraTS20 UAD | DICE: 39.45 -> 52.66 | Improved |
| raise FRM threshold above 0.10 (ablation) | FDP + LDM · BraTS20 UAD | DICE: 52.66 -> 45.24 (0.15) -> 41.82 (0.30) | Hurt |
| increase prior context count from 128 to 256 (ablation) | FDP + LDM · BraTS20 UAD | DICE: 52.66 -> 52.74 | No clear effect |
| increase prior context count from 16 to 128 (ablation) | FDP + LDM · BraTS20 UAD | DICE: 45.94 -> 52.66 | Improved |
Closest papers by meaning
- Unraveling Normal Anatomy via Fluid-Driven Anomaly Randomization (CVPR 2025 · similarity 0.89)
- Few-Shot Domain-Adaptive Anomaly Detection for Cross-Site Brain Images (TPAMI 2021 · similarity 0.87)
- NOVA: A Benchmark for Rare Anomaly Localization and Clinical Reasoning in Brain MRI (NeurIPS 2025 · similarity 0.84)
- Correcting Deviations from Normality: A Reformulated Diffusion Model for Multi-Class Unsupervised Anomaly Detection (CVPR 2025 · similarity 0.83)
- Beyond Single-Modal Boundary: Cross-Modal Anomaly Detection through Visual Prototype and Harmonization (CVPR 2025 · similarity 0.82)
- Motion Artifact Removal in Pixel-Frequency Domain via Alternate Masks and Diffusion Model (AAAI 2025 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
